How to create a data-driven culture – and why it matters

After talking to Kitty Kolding, CEO of Chrysalis Partners, it’s obvious that the companies that really respect the data stand to derive the most value from it.

transformation / conversion / data cubes shift from one color to another
Guirong Hao / Valery Brozhinsky / Getty Images
Current Job Listings

CIOs, CDOs and CAOs across industries wrestle with establishing a data driven culture and maximizing the value from data while managing the risk. In a recent Harvard Business Review article, Randy Bean and Thomas Davenport reported alarming findings from the NewVantage Partners’ 2019 Big Data and AI Executive Survey of about 64 C-level technology and business executives. They reported that:

  • 72% of survey participants report that they have yet to forge a data culture
  • 69% report that they have not created a data-driven organization
  • 53% state that they are not yet treating data as a business asset
  • 52% admit that they are not competing on data and analytics.
kitty kolding Chrysalis Partners

Kitty Kolding, CEO of Chrysalis Partners

Clearly leading corporations are struggling to become data-driven. In a wide-ranging interview with Kitty Kolding, CEO of Chrysalis Partners, a thought leader and a bold innovator in data monetization, appraisal and articulation, I explored how companies can get better at:

  • Elevating data as an enterprise asset through data monetization
  • Communicating the value of and the value from data
  • Driving towards a data driven culture

Can you describe your data and analytics journey culminating in the founding of Chrysalis Partners?

My journey with data started with my very first job out of college as a commercial real estate appraiser in Chicago, where I saw that there was a real paucity of transactional data for commercial comparables. I started a data company to fix that issue which took me on an 8-year journey which was so much fun but incredibly demanding.

I became intrigued with perfecting inefficient ecosystems where data is missing and where processes are inefficient, clunky or not automated. So, I sold my data company and in the early 1990s moved to the Bay area, the epicenter of the early internet ventures. I ended up working for a company in the syndicated market research space helping companies with internet-based business models. We did some fascinating work of collecting data, putting them in an analytic construct, then through prescription and prediction support the transition of old business models to new internet-based business models.

From there I went to run an experiential marketing company which tracked how consumers were interacting with brands and their reactions to new product launches. After that I led a company called InfoCore which focused on data acquisition. We worked on data acquisition projects all over the world and through interactions with data suppliers across 95 countries gained a strong understanding of the value of data and the challenges and constraints in collecting them. We also worked from the buyers’ side serving as an intermediary, gaining great insights into the complexity of the data interchange – the difference in expectations, the complexities of data representation and data contract negotiations, difficulties in expressing the value of data and finally challenges of cross-cultural and cross-geographical business deals. I saw first-hand the inefficiencies, disconnections and missed opportunities.

So, we started Chrysalis with the idea of helping companies get as much value and revenue possible from their data. At Chrysalis, we work with a range of companies and we work with a range of data types from medical, property insurance, ad metrics, marketing, location and company’s internal data. I find the dynamics, business considerations and principles around data valuation, sharing, monetizing and communication to be common across industry and data types.

What is data monetization? Why is it relevant for companies for whom data is not a primary product or service?  

When people hear the term data monetization, they mistakenly assume it to mean exclusively as a cash register – a direct collection of revenue from data through collecting, formatting, synthesizing and productizing data. This is a great way to use data if that is your core business model. But I think the idea of data monetization is more about the process and rigor of monetizing data and connecting it to value. Data monetization is about forcing your processes to become more evidence driven. It is about making sure that you use the monetized data in your normal business process. 

So, at Chrysalis we help companies with speaking about their performance and results using data in a way that is beyond the typical sales or marketing or PR tools. We follow a very analytical approach to looking at data differently, repackaging it with appropriate granularity and specificity, and make the data come to life in a compelling and meaningful way. For example, we help companies to provision data in their Business Intelligence (BI) and analytic tools in an intelligent way highlighting challenges and opportunities and connecting the dots between problems solutions.

We often see companies pushing this idea of democratizing of data. While it is a worthy pursuit, you can derive greater value and probably reduce cost even more by focusing on monetizing the data through data driven decision making, customer interaction management and optimization of operations. Thus, linking data to organizational value, institutional value and client way helps you push data monetization beyond the simplistic idea of a data cash register.

What is data appraisal? What is the methodology and how should one think about applying it?

Data appraisal is a personal fixation of mine. I mentioned before that in my early career I was a commercial real estate appraiser. Fast forward to now I think the same concepts apply to the world of data. I believe that data asset should be on the corporate balance sheet as it has tangible value and it is expensive to gather and maintain. But data is not on the corporate balance sheets because of a lack of standardized data appraisal methodologies to empirically estimate the value of a given data asset.

I think this is solvable. There are analogs in the appraisal of commercial real estate, gemstones and art which are useful in considering a methodology for data appraisal. Just like any other asset I believe that value of data will and should appreciate or depreciate over time, depending on a variety of factors internal and external, macroeconomic and industry specific.

While there is no standardized methodology for data appraisal today, I think it is can be developed. I admit that is an audacious undertaking. I spent the last year or so creating a set of standards and approaches for valuing data assets. We created a simple data monetization scorecard to explain this concept. Once you enter the details of a data asset it considers a variety of factors and gives you a point of view of value of the data asset.

I am also in the process of founding the Data Appraisal Institute, which is a non-profit entity to popularize this methodology, train and certify people to be data appraisers and help companies through the process of data appraisal. It is early days, but we feel that this data appraisal practice and methodology can help companies establish numerical value for the data assets supported by a rigorous and empirical appraisal and valuation methodology. 

In my own experiences and from talking to peers I have seen mixed results from attempts to define the value of and the value from data. In your opinion what are the best practices for thinking about the notion of data value and communicating the same to enterprise leaders?

I think that is such an important topic. I admire how hard CDOs and CAOs work to establish a common vocabulary across the enterprise so that everyone is thinking about data the same way. But I wonder whether they are thinking enough about turning that around to be in a language of the constituency with whom they are interacting.

The business and functional leaders are used to doing things in a particular way which brought them success historically. They usually think about the dynamics, the inputs and outputs of the business using their own terms and definitions. One of the biggest stumbling blocks that we discovered as we rolled out our data centric services is the communication around data or as we call it “data articulation.” This is the proficiency of speaking to a business leader about data in that business leader’s language. It is the ability to make data digestible and actionable based on business objectives contextualized in the business and market realities of the world we live in. Sales people talk about things differently than those in operations or those in other parts of the business.

So, it is critical to understand this aspect and articulate data in the right business vocabulary and the right business context. This is not easy. I see more data and analytic initiatives failing because of a communication miss than an execution miss. Often you find in companies that the data is normalized, it is put into a beautiful system enabled with sophisticated BI and analytic tools. But that is not enough. You still have to get the data and the findings across to a human being which is a communication task.

This is our emphatic point of view substantiated by the evidence we see with clients over and over. Companies put a lot of money and effort into data and technology but not enough on what we call “humanizing data”. I am convinced that this is the missing dark matter of the analytic universe.

A persistent them in my conversations with data and analytic leaders is the “last mile problem.” While there is a huge amount of investment going towards data platforms, analytic platforms, teams and talent, the utilization and consumption of the same is often less than optimal. What are your experiences with the last mile problem of analytics and what are some considerations for solving this?

At Chrysalis we observe the same issue. When we built Chrysalis, we endeavored to better communicate data and the findings from data. As we started working with clients, we kept running into the same issue of underutilization of expensive thoughtfully implemented BI and analytic platforms. Our diagnosis of this is that it is often a communication issue. Often there is nothing wrong with the tools and they usually do exactly what you expected them to do. So, we started testing how we can extract the findings from these dashboards, humanize them by turning them into words and written narratives focused on the business user. These narratives helped to explain to the business how the data could help in doing their work better or to achieve their numbers or to reduce cost.

This prescriptive approach can be a couple of paragraphs which really helps explain the relevancy of the data and findings to the business user using business friendly terminology and context. We find that these narratives stimulate- the business users to go back to the data and analytic tools to explore further and start using the full range of capabilities as they were designed. We create “data humanization containers” – some short to fit on your mobile phone and some which are more detailed with images and narratives.

I don’t mean to suggest that a better data articulation is the only obstacle for data and analytics adoption. But it is a common but solvable one. To do this effectively we bring together a combination of talent – researchers, analysts, data engineers, creative writers, editors, designers, visualizers and graphic artists.

What are the best practices for enterprises to become more data driven?

I don’t think of data driven culture as a thing that you accomplish or a capability that you plug in. The organizations that do this the best are those who have decided that it requires what we call a daily devotion.

Deciding to become a data driven culture is like deciding to get a puppy. You don’t acquire the puppy and leave it at that. You care for it, you feed it and give it water, give it exercise, you monitor its health and create a relationship with that creature, and you make it part of your family. It is an everyday multiple times a day devotion and commitment.

So too is the data driven culture. You have to look at it as a continual evolution. It requires everyone in the organization saying that before I make a decision or take action, I am going to look at the data first. It is about people in the organization, with regular devotion, checking on the health of the data and incorporating it in the different decision-making processes. It is about the organization rewarding the people for doing that and calling attention to this behavior in a positive way from the top down.

Thus, companies that really respect the data stands to derive the most value from it. If you see evidence of that then you know that your company is on the right track towards becoming a data driven culture. But there are always roadblocks to this evolution. It could be a group of people or a department who says that our experience and our insight is better than any data. But the truth is if you are able to bring experience and data together, you will be a world-beater.

This article is published as part of the IDG Contributor Network. Want to Join?

Copyright © 2019 IDG Communications, Inc.

How do you compare to your peers? Find out in our 2019 State of the CIO report